Sunday, January 8, 2012

Pathology of Stabilisation in Complex Adaptive Systems


The core insight  of the resilience-stability tradeoff is that stability leads to loss of resilience. Therefore stabilisation too leads to increased systemic fragility. But there is a lot more to it. In comparing economic crises to forest fires and river floods, I have highlighted the common patterns to the process of system fragilisation which eventually leaves the system “manager” in a situation where there are no good options left.
Drawing upon the work of Mancur Olson, I have explored how the buildup of special interests means that stability is self-reinforcing. Once rent-seeking has achieved sufficient scale, “distributional coalitions have the incentive and..the power to prevent changes that would deprive them of their enlarged share of the social output”. But what if we “solve” the Olsonian problem? Would that mitigate the problem of increased stabilisation and fragility? In this post, I will argue that the cycle of fragility and collapse has much deeper roots than any particular form of democracy.
Read it at Macroeconomic Resilience
The Pathology of Stabilisation in Complex Adaptive Systems
by Ashwin

UPDATE: Steve Roth engages Ashwin in the comments:


. Damned interesting thinking. Wondering if you’ve read Dirk Bezemer’s work:
http://www.rug.nl/staff/d.j.bezemer/research
Tried a google search and didn’t find any references herein, thought you’d find it interesting.


. Steve Roth
18 Dec 11 at 3:19 pm



. 
The passage you quote is just a restatement of Minsky who is a Post-Keynesian. I am not arguing for avoiding inflation/deflation per se – it is the rapid cycling that is clearly a pathological phenomenon and a sign of fragility.
Of the Austrians, I have only really read Hayek whom I am significantly influenced by. But I’d be shocked if any Austrian has ever made this point. Even Minsky’s point is one that his centrist followers usually gloss over primarily because his prescription for avoiding this problem of rapid cycling is the ‘socialisation of investment’ just like Keynes.
My departure from Minsky is that resilience usually requires that we allow the system to fight off small sicknesses and restrict our interventions to fighting off severe sicknesses. This is slightly Schumpeterian but he would have probably said don’t intervene at all.
But my deeper point in this post is that there are deep-seated reasons for why we, at least in the West, choose stability. The current fragility only reflects the fact that we have gotten a lot better at the task of stabilisation compared to say a 100 years ago.
Even worse, in an uncertain world it’s hard to distinguish between stabilisation/fragilisation and more genuine progress. The two may even be inseparable.


. Ashwin
18 Dec 11 at 4:02 pm



. 
Steve – Thanks. Yes – I’ve read some of his papers such as this onehttp://www.levyinstitute.org/pubs/wp_665.pdf which is excellent.
There’s a reasonable amount of literature that treats the economy as a complex, non-linear system but there is precious little that treats it as an complex adaptive system. The reason, as Bezemer notes at the end of the paper, is that the model without micro-foundations is hard enough but an agent-based micro-founded model that also gets the disequilibrium dynamics of bank credit creation right is a very complex exercise. It is this adaptive, co-evolutionary dynamic of the system that I am most interested in.


. Ashwin
18 Dec 11 at 5:52 pm



. 
Do you follow Steve Keen’s modeling work? He definitely seems to be moving in the right direction.
When do you think we’ll see anything of interest out of the Santa Fe Institute guys?


. Steve Roth
18 Dec 11 at 6:47 pm


Steve – Yes, I have read Steve Keen’s work. Again, I like a lot of it. But his models are also focused on aggregates. If you ignore the adaptive consequences of your interventions, then you almost always end up with a pro-stabilisation conclusion.

I follow the work at Santa Fe quite keenly but in all honesty, I haven’t seen that much yet on the macro front that excites me.
Ashwin
18 Dec 11 at 9:14 pm

Anders engages Ashwin:


.
. Ashwin – I feel a similar discomfort to LH. I like your piece; but doesn’t it justify an entirely laissez-faire approach to macroeconomics? After all, automatic stabilisers are what they say – an attempt to stabilise the economy.
I appreciate you are not a gold bug, but your piece sounds pretty Schumpeterian – let natural ‘creative destruction’ take its course. You have commented before that you favour a dynamic form of capitalism where business failure is welcomed, but your piece seems to do more than justify letting businesses fail – it justifies poverty and growing inequality.


. Anders
20 Dec 11 at 11:05 am



. 
Anders – Quoting from what I said in my earlier post on river flood management and resilience, “Economic policy must allow the “river” of the macroeconomy to flow in a natural manner and restrict its interventions to insuring individual economic agents against the occasional severe flood.”


So long as automatic stabilisers only protect against severe disturbances, e.g. low levels of unemployment insurance, they have very few fragilising consequences.
Allowing business failure, removing barriers to entry, removing all rents that flow to businesses – these are critical. Insurance for individuals is not the issue here.
As I have illustrated on multiple occasions, it is these rents and implicit protections of regimes such as the Greenspan Put that are responsible for increased inequality. If we allow creative destruction to take place and protect the weak against the collateral damage, inequality will fall dramatically in an instant.


. Ashwin
20 Dec 11 at 1:07 pm



. 
Ashwin – I accept the counterfactual that if there had been better institutional protections erected over the last 100-200 years to banish rent-seeking and barriers to entry, there would be less inequality. But you are making a more ambitious claim: that despite the fortunes and wealth inequality (not to mention the Bourdieusian social and cultural capital) already in place today, somehow the creative destruction of certain business models would lead to inequality being reduced. This doesn’t seem so plausible; perhaps there is a particular post you can direct me to where you make this argument? Doesn’t this even go beyond what Olson was arguing?


To the rest of your comment, it seems you are saying that stabilisation of a complex system doesn’t *always* compromise resilience, but only does so where stabilisation is aimed at non-severe disturbances to the system. The distinction of ‘degree of severity of disturbance’ has a plausibility when it comes to flooding; but in macroeconomic terms, your account needs to characterise a business failure as ‘non-severe’, whilst characterising a person being genuinely without a job as a severe disturbance. I’m a progressive and so have sympathy with this, but it doesn’t seem an entirely obvious delineation that the average WSJ reader would accept.


. Anders
20 Dec 11 at 5:03 pm


.
.
. Anders – You could try the posts under the ‘inequality’ category such ashttp://www.macroresilience.com/2010/09/23/inequality-and-moral-hazard-rents-in-the-financial-sector/http://www.macroresilience.com/2011/11/07/rent-seeking-the-progressive-agenda-and-cash-transfers/ http://www.macroresilience.com/2011/06/13/the-influence-of-special-interests-and-rentiers-on-monetary-and-fiscal-policy/ and if you want an overarching macro framework under which I think, this posthttp://www.macroresilience.com/2011/11/02/innovation-stagnation-and-unemployment/ has it. On a simple level, simply removing this obsessive focus on asset price stabilisation itself would reduce inequality dramatically.


On focusing on individuals, the rationale isn’t really on non-severe vs severe disturbances but on the inability for individuals to game the system to as great a degree as limited-liability entities which I mentioned herehttp://www.macroresilience.com/2011/10/05/a-simple-policy-program-for-macroeconomic-resilience/ :


“In order to promote system resilience and minimise moral hazard, any system of direct transfers must be directed only at individuals and it must be a discretionary policy tool utilised only to mitigate against the risk of systemic crises. The discretionary element is crucial as tail risk protection directed at individuals has minimal moral hazard implications if it is uncertain even to the slightest degree. Transfers must not be directed to corporate entities – even uncertain tail-risk protection provided to corporates will eventually be gamed. The critical difference between individuals and corporates in this regard is the ability of stockholders and creditors to spread their bets across corporate entities and ensure that failure of any one bet has only a limited impact on the individual investors’ finances. In an individual’s case, the risk of failure is by definition concentrated and the uncertain nature of the transfer will ensure that moral hazard implications are minimal. ”


. Ashwin
21 Dec 11 at 1:13 pm


.


No comments: